I had a great time at this workshop last week. It was held at McGill University in Montréal, and there was an interesting mix of systems folks and theory folks. In a panel on the last day, there was a discussion about what is more important in the study of computer and complex networks today-- theory or practice?
The answer (from Matt Roughan) was that clearly both are important, but that the research that can be applied to practical problems will win. In a way this is a universal truth-- Newton, Gauss, Einstein, Daubechies all won because their theory was applicable. (Though I would like to refer you to my earlier post on Stigler's law.) On the other hand, the work of abstract mathematicians, a few of whom I know and love, won't be useful for at least another century... and without them we would be hosed (to use what I understand is a Canadian phrase).
It brought to mind a discussion I had heard some time ago about how in every discipline, either applications are leading theory or theory is leading applications. The speaker at that time said that he believed disciplines go through phases where either one or the other is ahead. I thought that was very interesting. His point was that in the study of sensor networks right now, the applications are leading the theory.